plot of chunk rt_rep
November 2025
Load libraries
library(ggplot2)
library("knitr")
knitr::opts_chunk$set(
echo = FALSE,
message = FALSE,
warning = FALSE,
fig.align='left',
fig.width=12,
fig.height=5,
fig.pos='H',
fig.path = "plots/",
out.width = "800px",
dev = c("png", "svg"),
dpi=500
)We also remove the condition WS (word salad) since for an error these items were present only on two of the four lists. The three levels of teh variable repetition (rep) and encoded as “read” (1st presentation), “rep1” (2nd presentation) and “rep2” (3rd presentation).
data = read.table('data/dataMaioli.tsv', sep='\t', header= TRUE, as.is=TRUE)
fac = c('scode','list','item','rep','POS','corrkey','correct','open')
for (f in fac) {
data[[f]]=factor(data [[f]])
}
levels(data$rep) = c('read','rep1','rep2')
print('*risposte corrette per soggetto*')## [1] "*risposte corrette per soggetto*"kable(table(data$scode, data$correct))| FALSE | TRUE | |
|---|---|---|
| 2NT | 3 | 858 |
| 58I | 5 | 856 |
| 6RK | 10 | 632 |
| ALD | 7 | 854 |
| BDG | 8 | 853 |
| BXF | 8 | 634 |
| CJT | 6 | 855 |
| EBQ | 5 | 637 |
| FGH | 6 | 855 |
| N8N | 8 | 853 |
| OKZ | 6 | 636 |
| PFG | 10 | 851 |
| QOL | 2 | 859 |
| RDU | 9 | 633 |
| SGC | 12 | 849 |
| SH2 | 3 | 639 |
| T4Y | 3 | 639 |
| XQR | 10 | 632 |
| YZ1 | 3 | 858 |
| ZMU | 8 | 634 |
print('*numero outlier per soggetto*')## [1] "*numero outlier per soggetto*"kable(table(data$scode[data$rt>3000]))| Var1 | Freq |
|---|---|
| 2NT | 2 |
| 58I | 4 |
| 6RK | 32 |
| ALD | 2 |
| BDG | 4 |
| BXF | 12 |
| CJT | 0 |
| EBQ | 1 |
| FGH | 1 |
| N8N | 0 |
| OKZ | 1 |
| PFG | 13 |
| QOL | 3 |
| RDU | 0 |
| SGC | 6 |
| SH2 | 14 |
| T4Y | 2 |
| XQR | 10 |
| YZ1 | 1 |
| ZMU | 1 |
d = subset(data, rt<=3000 & correct == TRUE & cond !='WS')
d$lrt = log(d$rt)
d$lff = log(d$countForm)
d$llf = log(d$countLemma)We first consider all regular sentences, pooling toghether AAT, SAND and the control condisions of the third block (SEC, SYC). These sentence have a variable length, syntactic structure and difficulty. Separate data sets for open and close words are also created.
reg = subset(d, cond %in% c('AAT','SAND','SEC','SYC'))
ow = subset(reg, open==1, select=-cwo)
cw = subset(reg, open!=1, select=-owo)These plots are on raw data to describe the data set and see if marcrosocpic trends emerge across trials and subjects.
ggplot(reg, aes(x=rep, y=rt))+ geom_violin() + geom_jitter(width=0.1, height=0, size=0.3)plot of chunk rt_rep
For this analisis word position is treated as a countinuos interval variable (it can be also considered, probably more correctly, as an ordinal factor).
RT vs word position
ggplot(reg, aes(x=wp, y=rt))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(reg$rep) +
ggtitle('RT as a function of word position') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_wp
RT vs word position split between open and close class word
ggplot(reg, aes(x=wp, y=rt, color=open))+ geom_jitter(width=0.2, height=0, size=0.1) + facet_wrap(reg$rep) +
ggtitle('RT as a function of word position (separate fit for open and close words)') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_wp2
Differently from previous plots we look as predictior the oreder of that type of word (open and close class) within the sentences.
Open words
ggplot(ow, aes(x=owo, y=rt))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(ow$rep) +
ggtitle('RT for open class words as a function their order in the sentence') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_wo_open
Close words
ggplot(cw, aes(x=cwo, y=rt))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(cw$rep) +
ggtitle('RT for close class words as a function their order in the sentence') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_wo_close
RT vs word length
ggplot(reg, aes(x=nchar, y=rt))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(reg$rep) +
ggtitle('RT as a function of number of characters') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_nchar
RT vs word length split for open and close class words
ggplot(reg, aes(x=nchar, y=rt, color=open))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(reg$rep) +
ggtitle('RT as a function of number of characters') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_nchar2
RT vs log word form frequency
ggplot(reg, aes(x=lff, y=rt))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(reg$rep) +
ggtitle('RT as a function of log of Frequency') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_lff
RT vs log word form frequency (open close words seprately)
ggplot(reg, aes(x=lff, y=rt, color=open))+ geom_jitter(width=0.2, height=0, size=0.3) + facet_wrap(reg$rep) +
ggtitle('RT as a function of log of Frequency') +
geom_smooth(method = "gam", formula = y ~ s(x, bs = "cs", fx = TRUE, k = 5))
plot of chunk rt_rep_lff2